Q1. Glacial Lakes

The data at this link contain labels of glacial lakes the Hindu Kush Himalaya, created during an ecological survey in 2015 by the International Centre for Integrated Mountain Development.

lakes = read_sf("../data/GL_3basins_2015.geojson")
num_lakes = length(unique(lakes$GL_ID))

largest_lakes = group_by(lakes, Sub_Basin) %>% 
  filter(Area == max(Area)) %>% 
  arrange(-Area) %>%
  select(GL_ID,Sub_Basin,Area,Latitude,Longitude)

tm_shape(largest_lakes %>% select(-GL_ID)) + 
  tm_polygons(col='Area', palette="Blues",legend.show=F) + 
  tm_facets(by='Sub_Basin', free.scales = F, ncol=5, scale.factor = 4) +
  tm_scale_bar()

lakes_subset = lakes %>% 
  filter((Latitude>=28.2 & Latitude<=28.4) & (Longitude>=85.8 & Longitude<=86.0))

basemap = cc_location(loc=c(85.9,28.3), buffer=15e3)

tm_shape(basemap) + 
  tm_rgb(alpha=0.9) + 
  tm_shape(lakes_subset) + 
  tm_polygons(col='deepskyblue3') + 
  tm_layout(bg.color=page_bg_color, inner.margins=c(0,0), frame=F, saturation=-1) + 
  tm_view(set.bounds = c(85.8,28.2,86.0,28.4), set.zoom.limits =c(11,14)) +
  tm_layout(title = "Lakes with latitude between 28.2 and 28.4 and with longitude between 85.8 and 86")

Q2. Australian Pharmaceuticals II

The PBS data set contains the number of orders filed every month for different classes of pharmaceutical drugs, as tracked by the Australian Pharmaceutical Benefits Scheme. The code below takes the full PBS data set and filters down to the 10 most commonly prescribed pharmaceutical types. This problem will ask you to implement and compare two approaches to visualizing this data set.

pbs_full <- read_csv("../data/PBS_random.csv") %>% 
  mutate(Month = as.Date(Month))

top_atcs <- pbs_full %>%
  group_by(ATC2_desc) %>%
  summarise(total = sum(Scripts)) %>%
  slice_max(total, n = 10) %>%
  pull(ATC2_desc)

pbs <- pbs_full %>%
  filter(ATC2_desc %in% top_atcs, Month > "2007-01-01")

ggplot(pbs) + 
  geom_area(aes(Month,Scripts/1e6, col=ATC2_desc, fill = ATC2_desc), alpha=0.6) +
  scale_x_date(expand = c(0, 0)) +
  scale_y_continuous(expand = c(0, 0), n.breaks=11) + 
  scale_fill_brewer(palette = "Paired") + 
  scale_colour_brewer(palette = "Paired") + 
  labs(
    title = "Sales of 10 most commonly prescribed pharma drugs from 2007 till date", 
    y = "Orders sold (in millions)", fill = "CLASS OF DRUG", col = "CLASS OF DRUG"
  )

ggplot(pbs) +
  geom_alluvium(aes(Month, Scripts/1e6, fill = ATC2_desc, col = ATC2_desc, alluvium = ATC2_desc), decreasing = FALSE, alpha = 0.6) +
  scale_x_date(expand = c(0, 0)) +
  scale_y_continuous(expand = c(0, 0), n.breaks=11) + 
  scale_fill_brewer(palette = "Paired") + 
  scale_colour_brewer(palette = "Paired") + 
  labs(
    title = "Sales of 10 most commonly prescribed pharma drugs from 2007 till date", 
    y = "Orders sold (in millions)", fill = "CLASS OF DRUG", col = "CLASS OF DRUG"
  )

Q3. Spotify Time Series II

The code below provides the number of Spotify streams for the 40 tracks with the highest stream count in the Spotify 100 dataset for 2017. This problem will ask you to explore a few different strategies that can be used to visualize this time series collection.

spotify_full <- read_csv("../data/spotify.csv")

top_songs <- spotify_full %>% 
  group_by(track_name) %>%
  summarise(total = sum(streams)) %>%
  slice_max(total, n = 40) %>%
  pull(track_name)

top_10_songs <- spotify_full %>% 
  group_by(track_name) %>%
  summarise(total = sum(streams)) %>%
  slice_max(total, n = 10) %>%
  pull(track_name)

spotify <- spotify_full %>% filter(track_name %in% top_songs)

# Data Preprocessing to cut down a long song name
spotify$track_name[spotify$track_name==unique(spotify$track_name)[14]] <- "I Don’t Wanna Live Forever"
spotify$track_name = trimws(str_replace(spotify$track_name, " \\(([^)]+)\\)", ""))

# TODO: Save plots as images and load them in the Knit output
ggplot(spotify, aes(x=date)) + 
  geom_line(aes(date, streams/1e6, group=track_name),col="forestgreen", alpha=0.5,size=0.6) +
  scale_x_date(expand = c(0,0)) +
  scale_y_continuous(expand = c(0,0), breaks = seq(0,12,1)) + 
  labs(title="Spotify's 40 top streamed songs in 2017", x="",y="Streams (in millions)") +
  theme(axis.ticks.x = element_line(), axis.text.x=element_text(size=7, hjust=0.7), panel.grid.minor = element_blank())

custom_greys_palette = colorRampPalette(c("black","#EAEAEA"), space = "Lab")
custom_greens_palette = colorRampPalette(c("#BAE4B3","forestgreen"), space = "Lab")
spotify_palette = c((custom_greys_palette(8)),custom_greens_palette(4))

# TODO: Save plots as images and load them in the Knit output
ggplot(spotify, aes(date,streams/1e6)) + 
  geom_horizon(aes(fill=..Cutpoints..), origin = 4, horizonscale = seq(0,12,1), alpha=0.8) + 
  facet_wrap(~ reorder(track_name, -streams), ncol=1, strip.position = 'left') +
  scale_x_date(expand = c(0,0)) + 
  scale_y_continuous(expand = c(0,0), breaks = seq(0,12,3)) + 
  scale_fill_manual(values = spotify_palette) + 
  labs(title="Spotify's top streamed songs in 2017", x="",y="",fill="Streams (in millions)") + 
  guides(fill = guide_legend(nrow=1, reverse=T)) +
  theme(
    axis.ticks.x = element_line(), 
    axis.text.x=element_text(hjust=0.9),
    axis.text.y=element_blank(), 
    strip.text.y.left = element_text(angle=0, hjust=0.5), 
    legend.position = "top")

Q4. CalFresh Enrollment II

In this problem, we will develop an interactively linked spatial and temporal visualization of enrollment data from CalFresh, a nutritional assistance program in California. We will use D3 in this problem.

Part (a)

Using a line generator, create a static line plot that visualizes change in enrollment for every county in California.

Part (b)

On the same page as part (a), create a choropleth map of California, shaded in by the average enrollment for that county over the full time window of the data set.

Q5. Temporal and Geospatial Commands

For each of the commands below, explain what role it serves in temporal or spatial data visualization. Describe a specific (if hypothetical) situation within which you could imagine using the function.

Q6. Visualization for Everyday Decisions

We routinely encounter spatial and temporal data in everyday life, from the from the dates on a concert calendar to the layout of buttons in an elevator. This problem asks you to critically reflect on the way these data are represented.

a. Describe one type of spatial or temporal data set (loosely defined) that you encounter in a nonacademic context. What visual encodings are used? Does it implement any interactivity?

I use the weather app on my phone every morning. The app shows a single-line plot of the temperature of the day every hour. It also shows a small icon next to every (hourly) point on the line to show if the sky is sunny, clear, cloudy, rainy, or snowy at the time-point, and another small water drop at the bottom showing the chances of precipitation in percentages. When I click on one time-point, the app shows a detailed description of the weather at the time.

b. What questions does the visualization help you answer? How easily can you arrive at an accurate answer?

I plan my outfits according to the results of this visualization everyday. If it is cold and snowing, I layer up and wear warm clothes. It is very easy to arrive at an accurate answer with this, as there is only one single line for the current day in this visualization, and only one point on the line for every hour.

c. In the spirit of the re-imagined parking sign project, propose an alternative representation of these data (or an alternative way of interacting with the same representation). Why did you propose the design that you did? What advantages do you think it has?

I had a fun thought on how this line plot can be shown differently. I would put up a circular clock and replace every hour point with the icons used to display the sky condition. I would fill up the circular clock 30% by water if chances of precipitation at that time is 30%. The hour hand would point towards the current hour and two small rectangular boxes inside the circle will show the numeric values of the temperature and the UV Index. I will swipe over the clock to see the same info for the next 12 hours.

One advantage it has is that I can see the weather trend of the day in a single glance. In the original line plot, I have to keep scrolling to the right to see the evening weather, given the limited screen-space of my mobile.